Build Redis cache configs with clustering, shard count, geo-replication, eviction policies, persistence, and patch schedules.
Build Redis cache configs with clustering, shard count, geo-replication, eviction policies, persistence, patch schedules, and private endpoints.
Required Fields
nameresourceGroupsku.namesku.capacityminimumTlsVersionredisConfigurationOutput will appear here...Build an Azure Cache for Redis Premium-tier configuration with clustering (shardCount partitioning data across nodes for horizontal scale), zone-redundant replication, and geo-replication linking a secondary cache in another region as a read-only replica for disaster recovery. Clustering and geo-replication are genuinely different scaling dimensions that are easy to conflate: clustering (shardCount) partitions the keyspace across multiple nodes within one region for more total memory/throughput, while geo-replication links a whole separate cache instance in another region for regional failover, enabling one doesn't substitute for the other, a clustered cache with no geo-replication has no protection against a regional outage, and geo-replication without clustering doesn't add any additional capacity within a single region.
No, geo-replication's purpose is regional disaster recovery, a read-only secondary cache instance in another region kept in sync with the primary, it doesn't add processing capacity or memory to the primary region's cache. For additional in-region capacity, clustering (increasing shardCount) is the relevant mechanism, the two features solve different problems and often need to be considered together, not as alternatives.
No, geo-replication in Azure Cache for Redis requires an explicit administrative action (or automation you build) to promote the secondary to become the new primary during a failover event, it's not an automatic, transparent failover the way some other managed services provide. Plan and test the actual failover procedure ahead of time rather than assuming geo-replication alone provides zero-touch regional resilience.
allkeys-lru governs in-memory eviction behavior when the cache approaches its memory limit (evicting least-recently-used keys regardless of TTL), which is independent of RDB/AOF persistence, which governs whether cache contents survive a restart at all. Together, they mean the cache evicts under memory pressure during normal operation but also has a way to recover its state (from the RDB/AOF backup) after a planned or unplanned restart, rather than restarting completely empty.
The builder validates that name, resourceGroup, sku.name, sku.capacity, minimumTlsVersion, and redisConfiguration all resolve before accepting the JSON as a valid combined cache configuration; it can't verify the referenced private endpoint subnet or geo-replication linked server actually exist and are correctly paired, those are only confirmed against the live Azure Cache for Redis resource provider.
Don't conflate clustering (in-region horizontal scale) with geo-replication (cross-region DR), they solve different problems and a production cache handling both meaningful load and requiring DR protection typically needs both configured together, not one as a substitute for the other.
Geo-replication failover is a manual/automated-by-you process, not automatic, have a tested runbook (or automation) ready for actually promoting the secondary before you need it during a real regional incident.
RDB/AOF backup settings add real overhead and storage cost, but without them a cache restart (planned maintenance, a failure) starts completely cold, weigh that tradeoff deliberately for a cache whose warm-up time matters to downstream applications.
Was this tool helpful?
Disclaimer: This tool runs entirely in your browser. No data is sent to our servers. Always verify outputs before using them in production. AWS, Azure, and GCP are trademarks of their respective owners.